Power System State Estimation Considering Measurements Correlation

Power system state estimiaton is a filter block between the whole network measurements and the energy management system(EMS),it can estimate voltage magnitudes and phasors of all nodes in the power systems based on redundancy measurements.EMS’s subsequent analysis and calculations rely on the accuracy of state estimation results.Since the concept of state estimation has been put forward,scholars have focused on how to improve its estimation accuracy for decades.Benefited from the development of phasor measurement technology,phasor measurement units(PMU)are employed widely in power systems.Compared with supervisory control and data acquisition(SCADA)measurements,PMU measurements have the advantages of high precision and fast sampling rate.Based on these advantages,hybrid measurements with PMU measurements state estimation has a higher precision than conventional state estimation.However,with the increasing complexity of the power systems,the more accuracy results of state estimator should be supplied for EMS.In engineering practice both SCADA and PMU measurements are correlated,but conventional state estimation methods usually suppose that measurement errors are independent and obey Gauss distribution,namely measurement correlations are neglected,it is clear that estimating results will be influenced.So,in this paper,a power system state estimation method considering both SCADA and PMU measurement correlations is proposed to improve state estimation accuracy.The main contents of this paper are as follows:1.Analyzed the measurement principles of SCADA and PMU measurement,and the reasons why there are correlations.2.A weighted least squares state estimation considering SCADA measurement dependencies method(DWLS-SE)is proposed.The basic ideas are as follows: firstly,unscented transformation(UT)is adopted to calculate the statistical correlations of SCADA measurement,and error covariance matrix is obtained which contains measurement correlations in off-diagonal elements.Secondly,the conventional WLS state estimation’s weighting matrix is replaced by the off-diagonal error covariance matrix to include measurement correlations into state estimation.Lastly,estimation results are obtained after run DWLS-SE method.To verify the validity of the proposed method,some simulations on different IEEE standard test systems are implemented,which shows that results of DWLS-SE method are superior to conventional WLS state estimation.3.Based on SCADA/PMU hybrid measurements,a hybrid measurement state estimation considering both SCADA and PMU measurement correlations method(DH-SE)is proposed.The basic ideas are as follows: firstly,for the reason that there are huge sampling rate differences between these two measurements,PMU buffer is proposed to remit the problem.Secondly,vector autoregressive model(VAR)is adopted to consider PMU measurements’ temporal and spatial correlation,after that,the forecasting PMU measurements and SCADA measurements are combined to implement DH-SE to get estimating results.Lastly,to verify the validity of the proposed method,some simulations on different IEEE standard test systems are implemented,which shows that results of proposed DH-SE method are superior to conventional hybrid measurement state estimation.